Functional Ecology 2016 doi: 10.1111/1365-2435.12728 THE ECOLOGY OF DE-EXTINCTION A mammoth undertaking: harnessing insight from functional ecology to shape de-extinction priority setting Douglas J. McCauley*,1, Molly Hardesty-Moore1, Benjamin S. Halpern2,3,4 and Hillary S. Young1 1 2 Department of Ecology, Evolution, and Marine Biology, University of California, Santa Barbara, CA 93106, USA; Bren 3 School of Environmental Science & Management, University of California, Santa Barbara, CA 93106, USA; National Center for Ecological Analysis and Synthesis, University of California, 735 State St. Suite 300, Santa Barbara, CA 93101, USA; and 4Imperial College London, Silwood Park Campus, Buckhurst Rd, Ascot SL57PY, UK Summary 1. De-extinction, or the process of resurrecting extinct species, has been advanced as a promis- ing new tool in conservation biology. Most scientific discussion of de-extinction has thus far focused on the methodology and ethics of bringing once-extinct species back to life. We ask: How can de-extinction be strategically shaped into a service that maximally benefits ecological communities and ecosystems? 2. Ecologists often indicate that the worst facet of extinction is the associated loss of ecological function. Several decades of research on defining, classifying and tracking changes in portfolios of ecological function have generated a rich repository of information that should be mined to help guide de-extinction towards a future where its products can meaningfully restore extinction-induced loss of function. 3. Classifications of ecological function remain more subjective than other biological tax- onomies. Yet, there is a clear consensus among ecologists that the functions of certain species are less ecologically redundant than others. The loss of such functionally unique species can have proximate and cascading effects on community and ecosystem functioning. 4. We review, from an ecologist’s vantage point, efforts underway to use de-extinction to res- urrect the woolly mammoth and the passenger pigeon. These iconic case studies illustrate the opportunities and challenges ahead for restoring ecological function using de-extinction. 5. There is great risk that de-extinction could limit itself to the fabrication of products that mimic the biology of extinct species, but fail to resurrect their ecology. We suggest three ways that de-extinction may more meaningfully restore the functioning of once-extinct species: (i) select target species from guilds with low functional redundancy; (ii) concentrate on species that went extinct recently rather than older extinctions; and (iii) only work with species that can be restored to levels of abundance that meaningfully restore ecological function. Key-words: cascading effect, conservation, de-extinction, diversity, extinction, functional ecol- ogy, functional redundancy, mammoth, passenger pigeon, restoration de-extinction science. De-extinction, as reviewed by Sed- Introduction don, Moehrenschlager & Ewen (2014), is the process of Historically, extinction was permanent in a way that few using new developments in ancient DNA sequencing and biological processes are. The binary finality of extinction, genetic engineering to resurrect once-extinct species. The however, is being challenged by fast-moving progress in pros and cons of de-extinction have been productively debated (Sherkow & Greely 2013; Shapiro 2015a). Here, *Corresponding author. E-mail: [email protected] we endorse neither side of this debate. We instead accept © 2016 The Authors. Functional Ecology © 2016 British Ecological Society 2 D. J. McCauley et al. the premise that the tools of de-extinction will likely be increasingly standardized, but the process of systematically used in one form or another in the near future to help defining traits for animals is still dynamically evolving manage the conservation of biodiversity. (Blaum et al. 2011). By far the largest number of species With a few exceptions (Seddon, Moehrenschlager & that have been submitted for consideration as targets for Ewen 2014; Shapiro 2015a), little attention has been given de-extinction are animal species (Fig. 1). This bias means to whether the products of de-extinction can resurrect the de-extinction scientists will need to grapple with an as yet ecological functions of extinct species. As the current path- rapidly evolving set of complex principles for defining fau- ways of de-extinction are not capable of producing exact nal function. replicas of extinct species, the primary focus of de-extinc- Understanding the relationship between species diver- tion efforts should be on the production of functionally sity and functional diversity is essential to understanding equivalent proxies (Shapiro 2016). Losses of function can how extinction erodes function and, importantly, if, when have far-reaching consequences for extant species, can and how de-extinction may best promote the recovery of realign the physical and biological functioning of whole such function. Across very large spatial scales, a strong ecosystems and can cascade across vast reaches of ecologi- positive, but saturating, relationship has often been cal time (Barnosky et al. 2015). To ensure de-extinction observed between species diversity and functional diver- does not simply manufacture functionally ineffectual sity across a wide range of taxa (Lamanna et al. 2014). eco-zombies, much more discussion is needed on how At smaller spatial scales, however, there is more variation de-extinction can meaningfully recover the functions in the relationship between species diversity and func- once delivered by extinct species. tional diversity: some systems lose functional diversity To consider if and how de-extinction can be used to fill much faster than species diversity (D’Agata et al. 2014), the functional holes punched in Anthropocene ecosystems, and others show no relationship or exhibit an inverse pat- we review how ecological functions and functional diver- tern (Villeger, Mason & Mouillot 2008; Edwards et al. sity are measured, consider how and where functional 2013). redundancy occurs and review the challenges and opportu- nities for recovering function in two rapidly advancing Functional uniqueness of species de-extinction case studies: woolly mammoths (Mammuthus primigenius) and passenger pigeons (Ectopistes migrato- Are all species created equal in regard to their contribution rius). Through this review, it becomes clear that de-extinc- to ecosystem functioning? Several studies have explored tion science requires deeper considerations of the the patterns and distribution of functional redundancy ecological uniqueness of target species, ecological hystere- sis and interactions between the resurrection of a species’ abundance and its ecological function. Ecological function defined Here, we define ‘ecological function’ as the diverse forms of influence an organism has on ecological dynamics (Til- man 2001). Early measures of diversity in ecological com- munities emphasized species diversity. Interest in measuring functional diversity emerged decades later (Hooper & Vitousek 1997; Tilman 2001). This alternative view of ecological diversity stressed the value of quantify- ing not only what kinds of organisms are present in an ecosystem, but also what kind of ecological tasks these organisms deliver and how functional loss influences dynamical properties like community stability (Wardle, Bonner & Barker 2000). Measuring ecological function and linkages to extinction Fig. 1. Taxonomic breakdown of species (animals subdivided by class) located in a literature review of named candidates for Plant functional trait taxonomies have a longer history of de-extinction (see Table S2 for full species list and references). study and are frequently characterized based on morpho- De-extinction considerations have disproportionately focused on logical and life-history attributes. Animal functional traits animals (solid colours) with emphasis on vertebrates. Plants (green, hashed colours), as yet, have been little acknowledged in are more often based on the trophic interactions and beha- the de-extinction literature. Numbers indicate the absolute number vioural traits (see Table S1, Supporting Information). of species. Red = Mammalia; dark orange = Aves; light orange Rubrics for defining plant functional traits are becoming = Amphibia; dark yellow = Insecta; light yellow = Gastropoda. © 2016 The Authors. Functional Ecology © 2016 British Ecological Society, Functional Ecology Making de-extinction ecologically meaningful 3 within different ecosystems (Micheli & Halpern 2005; trophic level (Vander Zanden et al. 2000; Tinker, Bentall Mouillot et al. 2014), with the common result that the & Estes 2008). amount of redundancy varies by functional group. It De-extinction operates by recovering individuals of seems, however, universally clear that some functional extinct taxa. Such individuals, however, may represent an groups are very species-depauperate, and consequently, ecologically idiosyncratic and constrained view of the his- many systems contain species which deliver unique and toric breadth of intraspecific variability. This science sug- unreplicated functions (Halpern & Floeter 2008; Hoey & gests that de-extinction may need to manage for Bellwood 2009; Mouillot et al. 2014). intraspecific bottlenecks in ecological function along with Determining where low levels of functional redundancy genetic bottlenecks (see Steeves et al. 2016). are situated taxonomically and geographically can help make